Recombinant eukaryotic viral vectors have become a preferred means of gene transfer for many researchers and clinicians. In vivo gene transfer is a strategy in which a nucleic acid, usually in the form of DNA, is administered to effect expression of the protein product of the transferred gene in a location that is beneficial to the recipient. The benefit can be the induction of an immune response to the gene product, i.e., vaccination, or modification of the genetic repertoire of target cells for therapeutic purposes. This can be accomplished efficiently using a recombinant adenoviral vector encoding a so-called “transgene.” Adenoviral vectors have advantages over other vectors commonly employed for gene transfer (e.g., retroviral vectors) since adenoviral vectors (i) can be produced in high titers (i.e., up to 1013 viral particles/ml); (ii) they efficiently transfer genes to nonreplicating as well as replicating cells; (iii) recombination is rare; (iv) there are no known associations of human malignancies with adenoviral infections despite common human infection with adenoviruses; (v) the adenoviral genome can be manipulated to accommodate foreign genes that range in size; (vi) an adenoviral vector does not insert its DNA into the chromosome of a cell, so its effect is impermanent and unlikely to interfere with the cell's normal function; and (vii) live adenovirus, having as an essential characteristic the ability to replicate, has been safely used as a human vaccine (Straus, In Adenoviruses, Pienan Press, New York, N.Y., 451-496 (1984); Horwitz et al., In Virology, 2nd Ed., Fields et al., eds., Raven Press, New York, N.Y., 1679-1721 (1990); Berkner, BioTechniques, 6: 616 (1988); Chanock et al., IAMA, 195: 151 (1966); HajAhmad et al., J. Virol., 57: 267 (1986); and Ballay et al., EMBO J, 4: 3861 (1985)). The human adenovirus is one of the most widely used recombinant viral vectors in current viral vectored vaccine and gene therapy protocols.
In terms of general structure, all adenoviruses examined to date are nonenveloped, regular icosahedrons of about 65 to 80 nanometers in diameter. Adenoviruses are comprised of linear, double-stranded DNA that is complexed with core proteins and surrounded by the adenoviral capsid. The proteins of the capsid are the targets of neutralizing antibodies and the different serotypes possess distinct amino acid sequences in the capsid proteins that are on the outside of the viral particle.
Adenoviruses belong to the family Adenoviridae, which is divided into five genera, Mastadenovirus, Atadenovirus, Siadenovirus, Aviadenovirus, and Ichtadenovirus. The adenoviruses in the genus Mastadenovirus infect mammals only and include the human, chimpanzee, and monkey adenoviruses.
Adenoviruses provide an elegant and efficient means of transferring transgenes into cells. However, one problem encountered with the use of adenoviral vectors for gene transfer in vivo is the presence of pre-existing immunity to adenovirus that was acquired by the recipient through natural exposure to the adenoviruses. Primarily, infection with adenovirus throughout life induces the generation of antibodies to antigenic epitopes on adenoviral capsid proteins. If sufficient in titer, the antibodies can limit the efficacy of the adenovirus gene transfer vector. In addition, the administration of an adenovirus vector can induce immunity; thus an adenovirus may not be used more than once as an effective gene transfer vehicle. For instance, animal studies demonstrate that intravenous or local administration (e.g., to the lung, heart or peritoneum) of an adenoviral type 2 or 5 gene transfer vector can result in the production of antibodies directed against the vector which prevent expression from the same serotype vector administered 1 to 2 weeks later (see, e.g., Yei et al., Gene Therapy, 1: 192-200 (1994); Zabner et al., Nat. Gen., 6: 75-83 (1994); Setoguchi et al., Am. J. Respir. Cell. Mol. Biol., 10: 369-377 (1994); Kass-Eisler et al., Gene Therapy, 1: 395-402 (1994); Kass-Eisler et a)., Gene Therapy, 3: 154-162 (1996)). This is a drawback in adenoviral-mediated gene transfer, since many uses of an adenoviral vector (e.g., for inducing or boosting the immune response to a pathogen or providing a second dose of a therapeutic) require repeat administration. The mechanism by which antibodies directed against an adenovirus are able to prevent or reduce expression of an adenoviral-encoded gene is unclear. However, the phenomenon is loosely referred to as “neutralization”, and the responsible antibodies are termed “neutralizing antibodies.” Thus, to take full advantage of adenovirus vectors for in vivo gene transfer, novel types of adenoviruses are needed that (1) are not susceptible to neutralization by antibodies directed against another type, and (2) are not susceptible to neutralization by antibodies commonly found in the human population.
There are many different adenoviruses isolated from a broad range of animal hosts and adenoviruses are named by host first isolated from. I-lost animals from which adenoviruses have been isolated include mammals, birds, snakes, frogs, and fish. The mammalian hosts include, among others, primates such as monkeys, humans, and chimpanzees.
Humans and chimpanzees are very closely related and are grouped together as hominids. In contrast, monkeys are not grouped with humans and chimpanzees because there is a significantly greater evolutionary distance between them. The monkeys diverged between 25 and 35 million years ago from the hominids, whereas humans and chimpanzees diverged only about 7 million years ago (Samonte and Eichler, Nature Reviews Genetics, 3: 65-72 (2002)). These similarities and differences between humans, chimpanzees, and monkeys are consistent with documented host range restrictions of adenoviruses.
Many different ways for host range restriction occur. For example, wild-type human adenoviruses do not grow productively on monkey cells. In monkey cells infected with wild-type human adenovirus, the viral early genes are properly expressed (Feldman et al., J Bacteriol, 91: 813-8 (1966); Van der Vliet and Levine, Nature, 246: 170-4 (1973)), and viral DNA replication occurs normally (Rapp et al., J Bacteriol, 92: 931-6 (1966); Reich et al., PNAS, 55: 336-41 (1966); Friedman et al., J. Virol, 5: 586-97 (1970)). However, the expression of several late viral proteins is reduced. The block to late gene expression appears to be due to abnormal production of the viral late mRNAs (Klessig and Anderson, J. Virol, 16: 1650-68 (1975)), and this block can be overcome by a single mutation of the adenovirus DNA Binding Protein (DBP) (Klessig and Grodzicker, Cell, 17: 957-66 (1979)). Human adenoviruses that contain this single mutation in the DBP grow productively on monkey cells, suggesting that the key to the monkey/human block is centered on the roles of the DBP during the life cycle of the adenovirus.
In contrast to the monkey/human block are the observations that adenoviruses isolated from chimpanzees do not have a restriction in human cells and can be propagated efficiently (W. P. Rowe et al., Proc. Soc. Exp. Biol. Med., 97(2): 465-470 (1958); W. D. Hillis et al., American Journal of Epidemiology, 90(4): 344-353 (1969); N. Rogers et al., Nature, 216: 446-449 (1967)). In particular, replication of some chimpanzee adenovirus isolates was found to be more efficient in human than in monkey cells (M. Basnight et al., American Journal of Epidemiology, 94(2):166-171 (1971)). Adenoviruses isolated from other great apes species, such as gorillas and bonobos, have also recently been shown to grow in human cells (S. Roy et al., PLoS Pathogens, 5(7): e1000503 (2009)). Wild-type chimpanzee adenovirus replication in human cells does not require the expression of human adenovirus complementing factors, since E1-expressing cell lines (e.g., human embryonic kidney 293 cells, human retina PER.C6 cells) and non-expressing cell lines (A549 human lung epithelial carcinoma cells) have been used for their propagation (U.S. Pat. No. 6,083,716; S. F. Farina et al., Journal of Virology, 74(23):11603-11613 (2001); S. Roy et al, Virology, 324: 361-372 (2004); S. Roy et al., Human Gene Therapy, 15: 519-530 (2004); E. Fattori et al, Gene Therapy, 13(14):1088-1096 (2006); J. Skog et al., Molecular Therapy, 15(12): 2140-2145 (2007); D. Peruzzi et al., Vaccine, 27(9): 1293-1300 (2009)). The absence of a replication block is consistent with the close evolutionary distance between the human and chimpanzee lineages, which diverged only about 7 million years ago (Samonte & Eichler, Nature Reviews Genetics, 3: 65-72 (2002)).
Consistent with the greater divergence of hosts, a host range restriction of monkey adenoviruses for growth on human cells has been described (Am. J. Hyg., 68: 31 (1958); Virology, 35: 248 (1968); Savitskaya et al., Doklady Biochemistry, 375: 242 (2000); Alstein et al., JVi, 2: 488 (1968); Genetika, 39(6): 725-31 (June 2003)), and it has been hypothesized that the determinants are partially E4 and possibly E2. Savitskaya et al., supra, reported there was no growth of the monkey adenovirus SV7(C8) (now known as SV16 (ICTV 8th Report, p. 220)) on human embryonic kidney (HEK) cell line 293. Thus, an E1 region from a human adenovirus was not sufficient to alleviate the block to replication. The virus could grow on HEK-293 cells with Ad5 E4 region inserted (VK-10-9 cells). However, the VK-10-9 cells provided only partial alleviation of the replication block since replication was 40-fold lower than on CV1 cells (continuous line of green monkey kidney). This showed there was still a block to monkey virus replication in VK-10-9 cells. The authors concluded that E4 expression was too low, based on E4 ORF3 protein level (Krougliak and Graham, Hum Gene Ther., 6: 1575 (1995)), and a virus specific product was probably required (Savitskaya et al., supra). The authors then proposed the product might be encoded by the E2A gene, though additional study would be needed to clarify the problem. Savitskaya et al., supra, also notes that low E4 expression could have been the cause or an additional factor from E2A would be required for complete release of the replication block, and additional study was needed to define the causes. Interestingly, although the level of E4 expression in the VK-10-9 cells was reported to he significantly lower than that during wild type Ad5 replication, it was high enough for replication of an E4-deleted human adenovirus type 5 virus to the same level as wild type human Ad5 in HEK-293 cells (Krougliak and Graham 1995), further suggesting that the expression level of E4 was not the complete explanation for the species-specific block. Therefore, while it was believed that more E4 expression and/or an E2A product were required, neither was required. Also, it is apparent that the Ad 5 E4 function required for virus growth is separate from that required to overcome host range restriction of monkey adenoviruses on human cells because the E4 requirement for growth is not the same for host range determination. Thus, Savitskaya et al., supra, demonstrates that adenovirus E1 and E4 regions are likely not sufficient for alleviating the species block, and that other regions, in particular that encoding the DBP (E2A), are important.
In addition, in another study, an adenovirus-adenovirus hybrid of human Ad2 and SA7(C8) was shown to be defective for replication, suggesting that human E1 and monkey E4 are not compatible and that human adenovirus E1 is not sufficient for overcoming the host range restriction, which is consistent with the above described results where human E1 expressed from the cell did not change host range (Alstein et al., JVi, 2: 488 (1968); Savitskaya et al., supra). A different adenovirus-adenovirus hybrid between Ad2 and SA7(C8) was generated by growth of the two viruses under selection conditions to prevent Ad2 propagation (Grinenko et al., Molecular Genetics, Microbiology and Virology, 5:25 (2004)). Growth and selection of the hybrid virus on human cells (HEK-293) yielded a defective virus that had incorporated only the L3 region of SA7(C8). The authors note that both Ad2 E4 and E2A (encoding the DNA binding protein) were present and intact in the defective hybrid, and state that the gene E4 and possibly E2A are involved in the determination of species-specific host range, consistent with the earlier conclusions that more than E4 was required for alleviating the host range restriction. These results showed that only 10% of the Ad2 genome could be removed in order for a monkey—human adenovirus hybrid to grow on human cells, leaving 90% of the Ad2 genome to contain host range determining factors. Therefore, this hybrid did not provide further delineation of human adenovirus products required for growth of monkey adenovirus on human cells. Taken together, these reports showed that E4 plays a role in host range determination but other adenovirus genes also play a role. Furthermore, the E4 region is comprised of at least five known protein products, and despite these studies, the component or components of E4 that may have been necessary for the partial alleviation of the host range block were not identified.
Thus, there remains a need for methods that can alleviate, and even remove, the species-specific block or host range restriction which prevents a monkey adenovirus from propagating or replicating efficiently on human cells. There also remains a need for adenoviruses and adenoviral vectors which are capable of circumventing the pre-existing immunity to adenovirus in humans, The invention provides such methods, adenoviruses and vectors, and methods of using the same,